Measurement of cerebral white matter perfusion using pseudocontinuous arterial spin labeling 3T magnetic resonance imaging--an experimental and theoretical investigation of feasibility.

• Main Authors: Wen-Chau Wu, Shu-Chi Lin, Danny J Wang, Kuan-Lin Chen, Ying-Ding Li
• Format: Article
• Language: English
• Published: Public Library of Science (PLoS) 2013-01-01
• Series: PLoS ONE
• Online Access: http://europepmc.org/articles/PMC3855805?pdf=render

PURPOSE: This study was aimed to experimentally and numerically investigate the feasibility of measuring cerebral white matter perfusion using pseudocontinuous arterial spin labeling (PCASL) 3T magnetic resonance imaging (MRI) at a relatively fine resolution to mitigate partial volume effect from gray matter. MATERIALS AND METHODS: The Institutional Research Ethics Committee approved this study. On a clinical 3T MR system, ten healthy volunteers (5 females, 5 males, age = 28 ± 3 years) were scanned after providing written informed consent. PCASL imaging was performed with varied combinations of labeling duration (τ = 1000, 1500, 2000, and 2500 ms) and post-labeling delay (PLD = 1000, 1400, 1800, and 2200 ms), at a spatial resolution (1.56 x 1.56 x 5 mm(3)) finer than commonly used (3.5 x 3.5 mm(2), 5-8 mm in thickness). Computer simulations were performed to calculate the achievable perfusion-weighted signal-to-noise ratio at varied τ, PLD, and transit delay. RESULTS: Based on experimental and numerical data, the optimal τ and PLD were found to be 2000 ms and 1500-1800 ms, respectively, yielding adequate SNR (~2) to support perfusion measurement in the majority (~60%) of white matter. The measurement variability was about 9% in a one-week interval. The measured white matter perfusion and perfusion ratio of gray matter to white matter were 15.8-27.5 ml/100ml/min and 1.8-4.0, respectively, depending on spatial resolution as well as the amount of deep white matter included. CONCLUSION: PCASL 3T MRI is able to measure perfusion in the majority of cerebral white matter at an adequate signal-to-noise ratio by using appropriate tagging duration and post-labeling delay. Although pixel-wise comparison may not be possible, region-of-interest based flow quantification is feasible.